VISION. 319 



rying this design into effect we have the choice of two me- 

 thods, both of which we End resorted to by nature under 

 different circumstances. 



The first method consists in providing for each of these 

 single rays a separate tube, with darkened sides, allowing 

 the ray which traverses it, and no other, to fall on its re- 

 spective point of the retina, which is to be applied at the 

 opposite end of the tube. The most^onvenient form to be 

 given to the surface of the retina, which is to be spread out 

 to receive the rays from all these tubes, 

 appears to be that of a convex hemi- 

 sphere; and the most eligible distribution 

 of the tubes is the placing them so as to 

 constitute diverging radii, perpendicular, 

 in every part, to the surface of the retina. 

 This arrangement will be understood by 

 reference to Fig. 406, which represents 

 a section of the whole organ: (T, T,) being 

 the tubes disposed in radii every where 

 perpendicular to the convex hemispheri- 

 cal surface of the retina (R.) Thus will an image be formed, 

 composed of the direct rays from each respective point of 

 the objects, to which the tubes are directed; and these points 

 of the image will have, among themselves, the same rela- 

 tive situation as the external objects, from which they ori- 

 ginally proceeded, and which they will accordingly faith- 

 fully represent. 



The second method, which is nearly the inverse of the 

 first, consists in admitting the rays through a small aperture 

 into a cavity, on the opposite and internal side of which the 

 retina is expanded, forming a concave, instead of a convex 

 hemispherical surface. The mode in which this arrange- 

 ment is calculated to answer the intended purpose will be 

 easily understood by conceiving a chamber (as represented 

 in Fig. 407,) into which no light is allowed to enter, except 

 what is admitted through a small hole in^a shutter, so as to 

 fall on the opposite side of the room. It is evident that 



